On January 22, 2016, about 1530 central standard time, a Piper PA-31 airplane, N997DN, was substantially damaged during a forced landing near Denton, Texas. The commercial rated pilot received minor injuries. The personal flight was being conducted under the provisions of 14 Code of Federal Regulations Part 91 without a flight plan. Visual meteorological conditions prevailed at the time of the accident. The local flight departed Denton Enterprise Airport (KDTO), Denton, Texas, about 1500.

According to the pilot, the accident flight was a routine maintenance flight following maintenance and a 2-year period where the airplane had not flown regularly.

Several days prior to the accident flight, the pilot conducted an extensive preflight of the airplane and supervised the fueling of the airplane. He stated that the inboard tanks were filled to 56 gallons and the outboard tanks had 10 gallons of fuel added . The nacelle tanks were empty.

Before conducting a flight in the airplane, the pilot performed a maintenance test run of the engines. He stated that he started and shut down the engines on the inboard tanks and conducted the taxi and engine run-ups on the outboard tanks. He estimated that he ran the engines for about 30 minutes. Aside from issues with the fuel boost pumps and the left and right fuel flow indicators, he noted no anomalies with the airframe, engines, or related systems during the taxi and engine run-up.

The airplane sat for three days between the maintenance test run and the accident flight. During the preflight inspection he noted a fuel spot on the floor under the right main fuel tank sump. Further examination revealed that the plug was not seated correctly. He estimated that less than a cup of fuel was lost during that time. For the accident flight the pilot estimated the fuel on board at takeoff was 120 gallons – no additional fuel was added prior to the flight, nor did he visually verify the fuel quantity as no one had flown the airplane or ran the engines since his last engine run three days prior.

The run-up of both engines prior to the flight revealed no anomalies. The pilot departed with the fuel selected to the inboard tanks and then once at altitude switched to the outboard tanks. The fuel flow meters were now working and when he noted the right engine flow start to drop he switched back to the inboard tanks. He climbed to 1,000 feet, checked multiple systems and then conducted a low approach to Bishop Field. He climbed back up to 2,000 feet and was conducting the cruise checklist when the right engine surged. The pilot noted the sound and behavior consistent with no fuel/fuel flow. The pilot checked the fuel selector valves and trouble shot the engine surge without resolve.

The pilot obtained clearance to enter a downwind for runway 36 at DTO at which time the right engine lost power. Before the pilot could declare an emergency and secure the right engine the left engine surged and lost power. During the forced landing to the field both wings and engine nacelles were substantially damaged.

According to the Federal Aviation Administration (FAA) inspectors who responded to the scene, they could smell fuel on scene but found no evidence of fuel. There was no fuel in the fuel tanks and there was no pooling of fuel outside of the airplane in the debris field or where the airplane came to rest. Only a quart of fuel was recovered from the left fuel tank by the airplane recovery team.

An examination of the airframe, wings, and fuel system was conducted under the auspices of the National Transportation Safety Board investigator-in-charge, inspectors from the FAA, and an investigator from Piper Aircraft. The examination revealed that the fuel tanks and fuel system had not been compromised during the accident. No mechanical anomalies were noted with the engine, airframe, or airframe fuel system that would have precluded normal operation.

The airplane was equipped with a J.P. Instruments EDM-790 that had the capability to monitor and record exhaust gas temperature, cylinder head temperature, oil pressure and temperature, manifold pressure, outside air temperature, turbine inlet temperature, engine rpm, compressor discharge temperature, fuel flow, and battery voltage. The unit contained non-volatile memory for data storage of the recorded parameters.

The recorder was in good condition and data were extracted normally. The EDM contained 2.7 hours of data and 15 power cycles. Both the engine run and the accident flight were captured on the EDM. The engine run captured 36 minutes of engine data. The accident flight captured 30 minutes of data. Based upon the calibrations set by the operator, in the EDM, total fuel consumption for the engine run was 5.6 gallons and total fuel consumption for the accident flight was 19.2 gallons.

Performance data from the airplane flight manual indicate that fuel consumption (at high cruise) to be 54 gallons per hour (gph) for both engines, and 28 gph at maximum endurance. Using data from the pilot's operating manual and airplane flight manual, investigators estimated the fuel burn for the maintenance test run, 3 days prior to the accident flight, to be between 10 and 18 gallons. Further, investigators estimated fuel burn for the accident flight to be between 27 and 50 gallons. Fuel receipts confirmed the addition of 99.36 gallons of fuel, prior to the engine maintenance test run, on January 18, 2016.